Multidimensional Data Analysis and Neural Network-Enhanced Integrated Decision Support System for Intelligent Decision-Making

To realize intelligent decision-making, this study presents a method integrating multidimensional data analysis and Online Analytical Processing (OLAP). This method operates by establishing structural equations for multidimensional data analysis, quantifying potential decision-making factors and incorporating implicit variables associated with legal risks and regulatory requirements. In the experiment, specific cases were employed to verify the support degree of the analysis results for legal risks and regulatory requirements involved in data sources. The experimental results indicated that data sources conforming to both factual scenarios and legal provisions achieved a support degree exceeding 0.9. Thus, whether from the perspective of natural persons’ subjective judgment or the quantitative output of the proposed method, the superiority and practicality of this method are fully demonstrated. Meanwhile, combined with data association rules, the study establishes transaction matrices and calculates the support degree of each transaction for decision analysis. Additionally, for spatiotemporal multidimensional data, this study further suggests constructing spatiotemporal cubes, thereby enabling spatiotemporal feature analysis of both continuous and discontinuous multidimensional data. This study also investigates the training approaches of Neural Network-based Decision Support Systems (DSS) and upgrades the DSS to an Integrated Decision Support System (IDSS) — a modification that enhances the system’s capability to handle incomplete input information. In the experiment focusing on neural network error occurrence, specific cases were used to demonstrate the misleading effects of certain erroneous output results, and it is clarified that when the error rate exceeds 20%, the system fails to provide effective support for decision-making. Therefore, improvements to the neural network training method are required, and the training sample set must incorporate the legal requirements and policy regulations applicable to the enterprise.